Wash wants to know the structure of the train's programming - its directives, its capabilities, and its limitations. For the most part, it'll be in the form of inquiry - he thinks the train is what his world refers to as a dumb AI (intelligent but limited, sometimes capable of forming a personality) and that will inform his approach - but he's also tech-savvy enough to help and possibly even troubleshoot if something goes wrong.

What he has going for him is that he has the memories of one of his world's foremost AI scientists in his head - it's all very world-specific knowledge, but there is a lot of it and it is in depth, and he'll be doing his best to apply it to this situation in the hopes it'll provide some insight. The downside of that is, delving into those memories messes with his sense of self (to the point of losing his sense of self, if temporarily), and he tries to avoid doing it too often or too deeply. He may well be willing to go deep if this project is looking like it'll produce results, but it'll leave him shaky and rattled afterwards, and he'll be mostly out of commission for the rest of the day.

(Red vs Blue is weird, y'all.)